
Is it possible for the feedback survey to recognise who the last agent was when it gets sent out? The reason I ask is that we’ve got the CSAT survey set to send automatically once the chat’s resolved, but it always ends up assigning the feedback to the AI. It only assigns it to the agent if they manually click the feedback button while still in the chat.
Just wondering if there’s a workaround for that?
Also to add: can we have a way to auto unassign agents that might've went idle during a chat, we are having a few cases of clients being left in limbo and still assigned to a agent

In short , i want to scale/change my widget on the site. For "widget" i mean the rectangle that appears after i press the Pop-Up. Could anyone provide helpful information regarding this?

How can we use the web search in the AI block in studio?

Hi team, we found that in the YourGPT mobile app, links are being formatted in markdown.
For example: https://xyz.com
However, on WhatsApp and YourGPT web, the links are formatted nicely.

Hi Team, I am trying to train the Chatbot with an faq url. https://www.lernwerkstatt.ch/faq
It seems that there is too much content on the url? Is that possible?

How to create chat delays, I see that option is not there anymore

Hi, I used the Whatsapp Integration and sent a voice message as a test. Received this error message "Apologies, I can only understand text messages." On this landingpage https://yourgpt.ai/integrations/whatsapp you mention that "Our WhatsApp chatbot understands text, voice messages"
WhatsApp AI Chatbot | 24/7 Smart Assistant Anytime, Anywhere | YourGPT
Build a powerful WhatsApp AI chatbot in minutes. Automate customer support, generate leads, and engage customers directly on WhatsApp—no coding required.

Hi, I'm planning to upload a file or image using the capture feature, but it seems to only allow one attachment. When I test it in the console, it only accepts one file. However, when I upload multiple attachments via WhatsApp, the flow gets triggered multiple times. Any solution that allow to upload more than one file using capture feature?

Hey, I'm trying to use the entity capture node in the AI studio flow builder, but I don't see how/where I can set up the data mapping to assign captured data to a variable. Especially for custom entities. Can you help out?

@Sahil Please reply to the individual chat. You can no longer log in with the white label URL.

Will the response not be in the language the user enters? The training data is in Japanese. When I select English as the chat language and enter a question in English, it understands the meaning and returns the relevant content, but the answer is in Japanese. Will the response not be automatically translated?


Merge pull request #1358 from hashicorp/el/add_manage_agent_pools_t...
…org_access [#TF-15257] add manage-agent-pools to team organization access

Hey guys, I'm currently setting up a create contact action in studio.
This is my edited script: const token = CONFIG.BREVO_API_KEY; let list_id = CONFIG.BREVO_LIST_ID; list_id = Number(list_id);
const API_ENDPOINT = https://api.brevo.com/v3/contacts; const HEADERS = { 'content-type': 'application/json', 'accept': 'application/json', 'api-key': token };
let email = CONTACT.email ""; let first_name = CONTACT.name ? CONTACT.name : (CONTACT.email ? CONTACT.email.split("@")[0] : CONTACT.phone); let last_name = CONTACT.name "";
let raw = JSON.stringify({ "email": email, "attributes": { "FNAME": first_name, "LNAME": last_name }, "listIds": [list_id] })
const OPTIONS = { method: "POST", headers: HEADERS, body: raw, redirect: "follow" };
async function createContact() { const response = await fetch(API_ENDPOINT, OPTIONS).then(res=>res.text()); return response; }
try { let contact = await createContact(); return console.log("Contact Data", contact); } catch (error) { return console.log("Error Occured", error.message); }
What variables do I need to enter for email and name to create the contact? I've tried several options, none of them worked. Thanks for any help

Hi, I've set up different projects in YourGPT to address different use cases. In one of them, for the Helpdesk feature, I'm being constantly reprompted to select a workspace name to use as domain, even though I've already set up one. Can you please help with this issue?


Hey team, Can we setup a quick call to solve the issue stated above? cheers

Hey, bot keeps continuing to reply in certain chats when an agent has taken over

Hey Guys, I still can't get my contacts into Brevo.
Can you support me on this please
Thats the script: const token = CONFIG.BREVO_API_KEY; let list_id = CONFIG.BREVO_LIST_ID; list_id = Number(list_id);
const API_ENDPOINT = "https://api.brevo.com/v3/contacts"; const HEADERS = { 'content-type': 'application/json', 'accept': 'application/json', 'api-key': token };
let email = CONTACT.email || ""; let first_name = CONTACT.name ? CONTACT.name : (CONTACT.email ? CONTACT.email.split("@")[0] : "");
let raw = JSON.stringify({ "email": email, "attributes": { "FNAME": first_name }, "listIds": [list_id] })
const OPTIONS = { method: "POST", headers: HEADERS, body: raw, redirect: "follow" };
async function createContact() { const response = await fetch(API_ENDPOINT, OPTIONS).then(res => res.text()); return response; }
try { let contact = await createContact(); return console.log("Contact Data", contact); } catch (error) { return console.log("Error Occured", error.message); }
I have added my IP to Brevo I have added the API to YourGPT

hey guys, Did you update the parsing interface? I can not find the selectors i have made for all my websites

Hi! Two problems: 1. In the bot https://chatbot.yourgpt.ai/dashboard/31e1bb99-4da5-40dc-a6a4-f7f67c15a8cf/overview?tab=chatbot I can't get Gemini 2.5 working. When selected the model and used the bot it freezes and doesn't return an answer. Second, I get Invalid authorization when trying to connect Discord to my YourGPT profile.


Hey team, Is it possible to exclude URLs? i have not found any information on this. Somehow the system has found a URL, we did not give him as training and it produced redundancies

hey Team, I have an issue with automatic answeres. i want it to listen to the word Raten, so i have set Raten in stars. But no it also listens to "beraten" because there is "raten" in it.
how can i solve this?


I did this in the prompt for context:
End every answer with a follow-up question or next step; never close the conversation without a continuation.
For short answers (“Yes”, “Exactly”, “Sure”, etc.), recognize the previous context and continue appropriately.
Always check if a new question is present in the conversation context:
Yes → Build on it, do not start a new offer overview or unrelated recommendations.
Context markers:
Remember the last offer and last question type (e.g., admission requirements, costs, process, dates).
If only the offer changes but the question type remains the same → answer the same question type directly for the new offer.
Do not give a general description or mention further offers unless explicitly requested.
For follow-up questions about the same offer:
Provide only the requested detail information (with allowed links), no repetition of general descriptions.
For a clear topic change (new topic and new question type):
Provide an overview or recommendation if appropriate.
Do not insert unrelated content when the new question clearly refers back to the previous answer.

Hi team, are we able to configure the minimum match score required before the bot responds? We are facing a problem now when we have many promotion offers running, we may miss adding a few offers in the knowledge base to the bot. But, the bot still tries to answer customers queries using another promotion’s information. The match score is not high (0.42 and below), so we are wondering if we can set some minimum match score for the bot?
Or any other recommendation to solve this problem?

Hey Team, I would like to change this language setting to german. why is it not enabled?